The basic mechanism behind the El Niño phenomena is fairly well understood. A brief explanation follows below. A somewhat expanded treatment of some of the processes involved can be found in Bakun (1996).
The trade winds blow, usually with remarkable regularity, from east to west in equatorial regions of the Pacific Ocean. Near the equator Coriolis effects are small and there is significant downwind surface transport. Thus the trade winds tend to push ocean surface water from the eastern equatorial ocean toward the western equatorial ocean, shallowing the layer above the thermocline toward the east and deepening it toward the west. As a result, the sea surface tends to have a slight upward tilt (Fig. 1) along the equator, from east to west and the thermocline has a correspondingly steeper downward tilt from the eastern to the western ocean. (This is due to the much smaller density difference between surface and sub-thermocline waters compared to that between water and air. Under typical non-El Niño conditions the sea level difference across the Pacific along the equator exceeds 0.5 m and the thermocline dips from a very shallow or non-existent mixed layer in the eastern Pacific to some 150 m in the western Pacific.).
| Fig. 1 - Walker circulation diagram, acting in equator’s plane. |
|
When the thermocline is appreciably deeper than the depth to which significant wind-driven transport persists, the upwelling that results from surface Ekman divergence may largely contain surface layer water rather than thermocline water. Thus the equatorial upwelling driven by the tradewinds tends to bring up larger amounts of cooler thermocline waters in the eastern ocean, and smaller amounts in the western ocean. Consequently, the sea surface temperatures in the eastern part of the equatorial Pacific Ocean are, under normal (i.e., non-El Niño) conditions, the coolest for the latitude in any ocean. Conversely, those in the western Pacific are the warmest of any large ocean area on earth (Rasmusson, 1985). The result is the existence of a relative heat source to the lower atmosphere in the western part of the Pacific, which results in a tendency for air to be rising in that part of the ocean. This is accompanied by sinking air over the cooler eastern Pacific. These regions of rising and sinking air are connected by the westward-blowing trade winds at lower levels and a balancing eastward wind flow at upper levels, forming a closed atmospheric circulation cell in the plane of the equator (Fig. 1). This circulation is called "the Walker circulation" (Bjerknes, 1969) after the great climatologist who first described the "Southern Oscillation" (Walker, 1924), which we now understand is caused by changes in the Walker Circulation.
When air rises, moisture condenses and releases large amounts of latent heat of evaporation. This further accelerates the linked processes of upward motion, condensation and participation, leading to the enormous thunder cloud "towers" and verdant rain forests of the Indonesian end of the equatorial Pacific. In contrast, the cool sinking air of the eastern ocean is relatively dry, having had the moisture squeezed from it in its earlier rise to high altitude. Accordingly, the Pacific Coast of equatorial South America is relatively arid compared to Indonesia, and also compared to the Brazilian rain forest on the other (Atlantic) side of the Andes mountain chain.
When the surface waters of the eastern Pacific warm (i.e., as a result of El Niño), the trans-ocean difference in oceanic heat supply to the atmosphere decreases. Thus the Walker circulation tends to slow down. Less rising air in the western ocean and less sinking air in the eastern ocean means that the atmospheric pressure difference between the eastern and western Pacific decreases. Interannual fluctuations of this pressure difference were recognized by Sir Gilbert Walker (1924) as being linked to anomalous weather patterns in India and other regions of the world. He named these fluctuations the Southern Oscillation. An index of the trans-ocean pressure difference, formed by taking differences in barometric pressure measured at stations on different sides of the tropical Pacific (e.g., pressure at Darwin, Australia, subtracted from that at Tahiti or Easter Island, etc.), is called the Southern Oscillation Index.
This mechanism is intrinsically unstable. When the eastern Pacific cools, the Walker circulation speeds up, increasing the trade winds. This increases the equatorial upwelling of cool water in the eastern part of the ocean, while driving the thermocline even deeper in the western part. This reinforces the east-west sea surface temperature contrast which in turn feeds back into a continually intensifying Walker circulation. Similarly, during a warm phase in the eastern Pacific, the Walker circulation tends to slow down. The equatorial upwelling relaxes and the east-west slope in the thermocline decreases. This decreases the sea surface temperature contrast, which feeds back into a further relaxed Walker circulation. Thus both cooling and warming phases in the eastern Pacific feature positive feedback loops, which tend to further intensify anomalous conditions. Accordingly, there is an intrinsic tendency for widely-differing "opposite states", rather than for a "central tendency" around which random fluctuations may be distributed.
Quinn (1974) noticed that El Niño episodes often occurred following the buildup of the Southern Oscillation index to particularly high levels, after which the index would drop precipitously (corresponding to the collapse of the Walker circulation in response to the episodic warming of the eastern Pacific that characterizes El Niño). Wyrtki (1975) then provided the scenario by which a pause in the trade wind circulation allows the "breaking free" of the sea surface "bulge" that the trade winds had pushed up in the western equatorial ocean. The bulge then travels eastward across the ocean by a process known as an "equatorial Kelvin wave" (e.g., see Bakun, 1996), deepening the thermocline as it progresses and suppressing the cooling action of the equatorial upwelling. This begins to shift the warm sea-surface conditions eastward, collapsing the Walker circulation and shifting the entire system into a warming phase. When the equatorial Kelvin wave reaches the continent of South America, it then becomes coastally-trapped and propagates poleward in both hemispheres, deepening the thermocline as it progresses.
The deepened thermocline along the coastal boundary produced by this "downwelling" Kelvin wave may result in the existence of anomalously warm, nutrient-deficient surface-layer water at the depths from which the coastal upwelling process draws its source waters (Wyrtki, 1975; Shkedy et al., 1994). This suppresses the surface cooling along the coast normally caused by upwelling, and also the associated high photosynthetic production. Moreover the sea surface bulge (or local pressure "High") which propagates along the coast would carry an imbedded poleward flow pattern, which would tend to directly advect conditions and organisms toward the poles and also serve to induce convergence of less productive offshore surface waters toward the coast in the vicinity of the passing wave front. These effects generally correspond to observed conditions that typify El Niño episodes off the west coasts of North and South America.
Originally, El Niño was thought to be a local phenomenon of the coast of Ecuador and Peru. Now that its Pacific-wide (and global) scale and consequences are recognized, the name "ENSO" (El Niño - Southern Oscillation) is often used to distinguish the Pacific-wide phenomena from its local South American consequences.